Epigenome editing of human hematopoietic stem cells enables sustained and reversible thrombosis prevention

This study demonstrates that transient delivery of DNA methylation-based epigenome editors to human hematopoietic stem cells enables stable, heritable, and reversible silencing of the platelet integrin ITGB3, resulting in sustained prevention of thrombosis through impaired platelet aggregation.

The Gist

Imagine your body's blood vessels as a busy highway system. Sometimes, this highway gets blocked by a traffic jam of blood cells called platelets. When these platelets clump together too easily, they form a clot (thrombosis), which can cause heart attacks or strokes.

Currently, the only way to keep this highway clear is to take a daily pill (like aspirin) that tells the platelets to "calm down." But this has two big problems:

1. You have to remember to take it every day. If you miss a dose, the traffic jam can happen.
2. It doesn't work for everyone. Some people's platelets are just too stubborn and ignore the pills.

The Big Idea: A "One-and-Done" Fix
This new study by Ye and colleagues proposes a revolutionary solution: instead of treating the traffic jam every day, let's reprogram the factory that makes the cars.

That factory is your Bone Marrow, and the workers are called Hematopoietic Stem Cells (HSCs). These are the "master cells" that constantly produce new blood cells, including platelets. If we can teach these master cells to make "calm" platelets permanently, you would only need the treatment once, and it would last a lifetime.

The Problem with Previous "Fixes"

Scientists have tried using gene editing (like CRISPR) to cut out the "bad instructions" in the DNA of these master cells. Think of this like taking a pair of scissors to a recipe book and cutting out a paragraph.

• The Risk: Cutting the book can leave "scars" (permanent damage to the DNA) or accidentally cut the wrong page, causing new problems.

The New Solution: The "Highlighter" Method (Epigenome Editing)

Instead of cutting the book, this team used a high-tech highlighter. They didn't change the text (the DNA sequence); they just highlighted the "stop" instructions so the cell's machinery ignores them.

They used a tool called CHARM.

• How it works: Imagine the DNA is a long string of beads. The CHARM tool puts a "Do Not Read" sticker (a chemical tag called methylation) on the specific bead that tells the cell to make the "sticky" part of the platelet.
• The Magic: When the cell divides to make new cells, it copies the beads and the "Do Not Read" stickers. So, every new platelet made from that factory is born with the "calm" instruction already written on it.

What They Did in the Lab

1. The Test: They picked a specific "sticky" part of the platelet called ITGB3. This is like the glue that makes platelets stick together to form a clot.
2. The Edit: They took stem cells from human donors, gave them the CHARM tool via a temporary RNA messenger (like a text message that disappears after being read), and told it to put "Do Not Read" stickers on the ITGB3 gene.
3. The Result:
   • Permanent: Even after the stem cells divided many times and turned into mature platelets, the "glue" was still missing. The platelets couldn't stick together.
   • Safe: They checked the rest of the "recipe book" and found no accidental cuts or scars. The cells were healthy and worked normally in other ways.
   • Reversible: Here is the safety net. If they ever needed to turn the "glue" back on (maybe if the patient needed to heal a wound quickly), they used a second tool (a "eraser") to wipe off the stickers. The gene started working again.

The "Traffic Jam" Test

They took these edited cells and grew them into platelets in a dish. When they tried to make these new platelets clump together (mimicking a clot), they failed. They were like cars that had lost their bumper stickers and couldn't stick to each other. The "traffic jam" never happened.

Why This Matters

• One-Time Treatment: You could get this treatment once, and your body would naturally produce "anti-clotting" platelets forever. No more daily pills.
• Safer: Because it doesn't cut the DNA, it leaves fewer scars and is potentially safer than current gene-editing methods.
• Versatile: They showed this works on other "sticky" genes too, meaning it could be customized for different types of clotting risks.

The Bottom Line

Think of this as teaching your body's factory to build "non-stick" platelets instead of "sticky" ones. It's a permanent, reversible, and safe way to prevent heart attacks and strokes without needing to remember a daily pill. While this is still in the research phase (tested in mice and lab dishes), it opens the door to a future where clotting diseases are cured with a single visit to the clinic.

Technical Summary

1. Problem Statement

Thrombosis (blood clot formation) is a leading cause of cardiovascular and cerebrovascular diseases. Current standard-of-care treatments involve lifelong daily administration of antiplatelet drugs (e.g., aspirin, clopidogrel). These therapies face significant limitations:

• Poor Compliance: Daily dosing is difficult for patients to maintain.
• Resistance: Some patients develop resistance or have persistently high platelet reactivity despite treatment.
• Genome Editing Risks: While gene editing (e.g., CRISPR-Cas9) offers a one-time cure for genetic disorders, it often leaves permanent "scars" in the genome (indels, off-target mutations, chromosomal rearrangements) and lacks reversibility.

The authors propose epigenome editing of Hematopoietic Stem Cells (HSCs) as a solution. By targeting the HSC compartment, a single intervention could durably reprogram the output of all downstream platelets. Crucially, epigenome editing modulates gene expression without altering the DNA sequence and offers the potential for reversibility.

2. Methodology

The study utilized CHARM (Coupled Histone tail for Autoinhibition Release of Methyltransferase), a DNA methylation-based epigenome editor.

• Delivery System: Transient delivery of mRNA encoding the CHARM construct (dCas9-KRAB-DNMT3A-DNMT3L) and a single-guide RNA (sgRNA) via electroporation into primary human CD34+ Hematopoietic Stem and Progenitor Cells (HSPCs).
• Target Selection:
  • Primary Target: ITGB3 (Integrin Beta-3), a subunit of the $\alpha_{IIb}\beta_3$ receptor essential for platelet aggregation.
  • Secondary Targets: GP1BB (GPIb-vWF adhesion), ANO6 (Phospholipid scramblase/PS exposure), and SERPINH1 (Collagen response).
  • Control Targets: CD47, CD9, B2M, and the safe-harbor locus AAVS1.
• Experimental Models:
  • In Vitro: Expansion of HSPCs, differentiation into megakaryocytes (MKs) and erythroid lineages, and serial colony replating assays to test self-renewal.
  • In Vivo: Xenotransplantation of edited human CD34+ cells into NBSGW immunodeficient mice to assess long-term engraftment and multilineage reconstitution.
  • Functional Assays: Flow cytometry for surface protein expression, Whole-Genome Bisulfite Sequencing (WGBS) for methylation mapping, RNA-seq for transcriptomics, and platelet aggregation assays (mixing culture-derived platelets with donor platelets).
• Reversibility Test: Sequential editing using CHARM for silencing followed by dCas9-TET1 (a demethylase) to erase methylation marks and restore gene expression.

3. Key Contributions

• First Demonstration of Durable Epigenetic Silencing in Human HSCs: The study proves that DNA methylation-based silencing can be established in HSCs and faithfully inherited through extensive self-renewal and differentiation, overcoming the challenge of epigenetic memory in rapidly dividing stem cells.
• Reversibility in HSCs: It demonstrates that epigenetic silencing in HSCs is not permanent; targeted demethylation can successfully reverse the silencing, restoring gene function.
• Therapeutic Platform for Thrombosis: It establishes a "one-time" strategy to prevent thrombosis by engineering HSCs to produce platelets with impaired aggregation capabilities, bypassing the need for daily medication.
• Safety and Specificity: The approach shows high specificity with minimal off-target effects on the methylome or transcriptome and does not compromise HSC viability or engraftment potential.

4. Key Results

A. Optimization and Durability of Silencing

• Comparison: CRISPRi (histone modification) silencing was transient and faded over days. In contrast, CHARM (DNA methylation) achieved robust (~80-95%) and sustained silencing of CD47 in HSPCs and their progeny.
• Differentiation: Silencing persisted through megakaryocytic and erythroid differentiation. ~95% of mature megakaryocytes and ~99% of erythroblasts maintained the silenced state.
• Self-Renewal: Silencing was maintained through:
  • Long-term cytokine-free culture (3 weeks).
  • Serial replating (up to tertiary colonies), indicating transmission through stem cell division.
  • In Vivo Engraftment: In NBSGW mice, edited HSCs engrafted robustly (comparable to controls) and maintained ITGB3 silencing in human CD34+ cells 16 weeks post-transplant.

B. Functional Impact on Platelets (ITGB3 Target)

• Gene Silencing: CHARM targeting of the ITGB3 promoter induced high levels of DNA methylation (~1.5 kb spread) and reduced ITGB3 mRNA by ~90%.
• Protein Loss: Mature megakaryocytes and derived platelets showed near-complete loss of CD61 (ITGB3 protein) surface expression.
• Aggregation Defect: Culture-derived platelets (CDPs) from edited cells failed to aggregate with donor platelets upon stimulation with ADP, TRAP, or U46619, mimicking the phenotype of Glanzmann thrombasthenia (a bleeding disorder caused by ITGB3 loss).
• Specificity: WGBS and RNA-seq revealed minimal off-target methylation or transcriptional changes. The only notable on-target effect was a slight upregulation of the neighboring gene MYL4, likely due to local chromatin remodeling, but this did not alter global programs.

C. Reversibility

• Sequential delivery of dCas9-TET1 to ITGB3-silenced HSCs successfully removed DNA methylation marks.
• This resulted in the restoration of ITGB3 mRNA and CD61 protein expression in differentiated megakaryocytes, proving the system is controllable and reversible.

D. Broad Applicability

The platform was successfully applied to other platelet targets:

• GP1BB: Silencing led to loss of CD42b and defective ristocetin-induced aggregation (mimicking Bernard-Soulier syndrome).
• ANO6: Silencing reduced phosphatidylserine (PS) exposure upon calcium stimulation, impairing procoagulant activity.
• SERPINH1: Silencing reduced collagen-induced aggregation.

5. Significance and Implications

• Therapeutic Paradigm Shift: This work proposes a shift from chronic pharmacological management of thrombosis to a one-time, cell-based curative intervention.
• Safety Profile: Unlike CRISPR-Cas9 genome editing, which risks permanent genomic damage (indels, chromothripsis), epigenome editing avoids DNA breaks. The reversibility feature adds a critical safety layer, allowing gene function to be restored if bleeding risks become too high.
• Versatility: The ability to target multiple genes involved in different stages of thrombus formation (adhesion, activation, aggregation) allows for customizable therapeutic profiles with varying bleeding risks.
• Future Outlook: While the study used ex vivo editing, the authors envision future in vivo delivery using lipid nanoparticles or virus-like particles. The study highlights that HSC epigenome editing is a viable, durable, and safe platform for treating not only thrombosis but potentially other hematopoietic disorders.

Limitations Noted:

• The NBSGW mouse model supports human HSC engraftment but produces very low levels of human platelets, preventing direct in vivo assessment of thrombosis prevention in a living organism.
• The approach is currently limited to genes with CpG-rich promoters; genes lacking these features or requiring enhancer targeting were not explored.